Edge Impulse with Mobile Phone Application Using Accelerometer

Introduction

As the mobile phone is compatible with Edge Impulse, it serves as the primary device for this project. The project utilizes an accelerometer, a device that measures motion or acceleration in a structure, as it is widely available on mobile phones. The goal is to measure the phone's motion and create a model that can identify these movements.

Hardware Requirement

• Mobile phone

Project Development

To create the project on Edge Impulse using a mobile phone device, it is crucial to consider seven necessary stages: 

1. Device

• Set up the mobile device by scanning the QR code to connect to the phone.

• Scan the QR code, and the website will guide you to the designated page. Ensure the message "Connected as..." is shown, confirming successful linkage between the mobile device and Edge Impulse.

2. Data Acquisition

• The sensor type of accelerometer is chosen from the list and the sample length can be adjusted on the same page.  

• When the “Start sampling” is clicked, the data for the motion of the mobile phone is recorded. For example, the motions of “up&down, shake, flip” are considered in this case.

• Modify the data type by changing its "label". Then, the data collected is uploaded and saved to the same data acquisition page.

3. Impulse Design

• The data can be graphically represented, whether in a chart or a table, but the impulse or feature must be produced and set up first by selecting them from the recommended list. For example, two block models must be set up: the processing block and the learning block. (The number of each block model relies on your requirements)

Note:  Remember to save the created impulse

• Each created impulse must be accessed and trained independently.

4. Retrain Model

• The model can undergo retraining with known parameters using the retrain model feature.

5. Live Classification

• The live classification category, also referred to as the test data stage, enables users to move the phone for collecting and classifying test data by clicking "Start sampling". Ensure to label each test data with the expected outcome.

6. Model Testing 

• Model testing helps represent the test data using graphical charts. After the testing, you can observe the output using the provided chart.

7. Deployment

• Scan the QR code to launch your building model on the mobile phone device.

• Move your mobile phone actively while sampling, and the result will be displayed on the same page.

Tutorial Video

This tutorial video clearly shows the entire project development.